Don't get me wrong, but metis is YOLO.

In 1932-33, Soviet collectivization destroyed local farming knowledge and produced a famine that killed somewhere between five and nine million people. It was one of the twentieth century’s great tragedies, and James Scott’s Seeing Like a State draws a straight line from the ideology that caused it — High Modernism, the belief that society can be rationally reorganized from above — to the disaster that followed.

But here’s a number that doesn’t appear in Scott’s book. Eight billion. That’s roughly how many people are alive today, most of them fed by the products of scientific agriculture. Synthetic fertilizers, high-yield crop varieties, mechanized farming. The Green Revolution, which saved millions from starvation in the second half of the twentieth century, was born from the same impulse as High Modernism: it is top-down, science-driven and generic, scaling standardized solutions across entire continents.

James Scott's Seeing Like a State is a brilliant book about the former kind of outcome. But it has little to say about the latter.1 This has allowed a generation of readers to walk away with a clean takeaway: Local knowledge good, central planning bad. But that is, at best, half of the story. The question that even Scott cannot avoid raising, but which he never truly answers is: if High Modernism is just hubris, why does it sometimes work so spectacularly?

***
Behold Machiavelli at his desk! He is doing wonders of diplomacy, navigating the chaos of the Italian Wars with exactly the kind of knowledge Scott calls metis, the very opposite of High Modernism: intimate familiarity with every actor on the stage, a gut feeling honed by many years of observing the treachery and contingency of Italian politics, tacit understanding of who can be trusted and up to what point, heuristics about when to flatter and when to threaten. He is steering Florence through a turbulent era that could destroy it, and he is doing it ad hoc, by instinct and by experience.

And sometimes he gets it wrong. He backs the republic when the Medici are about to return. He bets on a citizen militia and gets beaten by Spanish professional army. His instincts are good, but metis is a gamble. It comes with no guarantees.

But later, he is doing something else. Sitting there by the candlelight, he is writing The Prince. He is trying to extract from his own hard-won intuitions, from his metis, a set of general principles, something that could be written down, taught, applied by someone who hasn’t spent years in the Florentine chancery. Whether he intends it or not, he is inventing modern political science.

***

Observe an SRE at work. Her service is down. The dashboards are red. 70% of user requests are failing. Every minute of downtime burns thousands of dollars. She glances at the logs, but there’s no time for a proper diagnosis. She decides to restart the database server. Will it help? Possibly. There’s a significant chance it won’t. There’s even a small likelihood it will make things worse. But years of experience make the calculus automatic. The risk is low, the expected payoff is positive. Restart!

Later that day, dashboards green, coffee in hand, she sits down to write the postmortem. A failed disk had sent the database into a busy loop, eating all the CPU. She proposes a fix: monitor disk health continuously and automatically restart the database if it fails. She assigns the ticket to a developer and will, in time, check that it was done. Because despite all her metis, all her hard-won ability to read a misbehaving system and make the right call under pressure, she does not want this to happen. She doesn’t want to be a hero. If she must make YOLO calls, so be it, but better to make them as rare as possible.

***

Watch a family of farmers, tending their small plot somewhere in the tropics. It is not a monoculture, not just wheat and nothing else. No, sir! That would be far too dangerous. If a drought hit or a blight spread, the family would have nothing to eat. Instead, the plot is a tangle of different crops, carefully arranged in a pattern that looks like chaos to a visiting agronomist but is anything but. Beans fix nitrogen in the soil. Tall stalks of maize shade the root vegetables below from the scorching sun. Squash covers the bare soil on the slope, keeping it from washing away when the rains come hard. A few fruit trees at the edges serve as windbreaks. Every combination has been tested by generations of failure and hunger, and the family is dripping in metis: they know exactly what goes where and why.

Impressive. But also precarious. Failure of one crop is survivable. Failure of two means children go hungry. Survival in August depends on whether you’ve read the sky correctly in March.

And when you ask the family what they want for their children, they won’t say “to inherit our skill, to farm as we do.” They would say: “a government job.”

***
Metis, Scott says, is a wide array of practical skills and acquired intelligence in responding to a constantly changing environment. Metis is like riding a bicycle. It can be learned, but not from a book. In a classroom you can teach the physics of it, but not the skill itself. Metis is what a master carpenter passes to his apprentice over years at the workbench. Metis is what firefighters, rescue squads, and paramedics draw on when dealing with emergencies. Metis is what a skilled diplomat, a cunning politician, and a general on the battlefield have in excess. Adapting quickly to the enemy's move and making the best out of limited resources. That is metis.

It is most needed when the situation is fluid, the future uncertain, and success depends not on general principles but on the particulars. When a freight ship enters a port, the captain steps back and hands the wheel to the local pilot. It is he who knows the unique features of local wind and wave patterns, the shifting sandbars, the unmarked reefs, the seasonal changes in microcurrents, the local traffic conditions and how to pilot in these waters at night. That pilot lives off metis.

Metis is indispensable when the environment is complex. It may not always be chaotic, but all too often it is. And where it is so, small causes cascade into large effects and make the future unpredictable. Not unpredictable because you are bad at prediction, but unpredictable in principle. There is no algorithm that will tell you the best course of action. No amount of data that will make the uncertainty go away. You have to guess. You have to feel your way through. You have to use metis.

Think of a magnetic pendulum, a pendulum hung above three magnets. Release it from a given position and it will swing back and forth, eventually coming to rest at one of the magnets. But which one? The answer depends, with high sensitivity, on where exactly you released it. Plot the outcomes by starting position and you get a fractal, a pattern where patches of solid color — starting points that all lead to the same magnet — are separated by regions where the tiniest shift sends you to a completely different outcome. Move the pendulum a hair’s breadth before you let go and it ends up at a different magnet. Replace your hand with a robot precise to a hair's breadth and it still won't suffice. The pattern is infinitely detailed. Zoom in arbitrarily and the boundary is just as tangled.

The real world, of course, is far worse than this toy example. Any system is constantly bombarded by random noise from the outside, affecting the outcome in unpredictable ways. But on the other hand, in reality, unlike in the experiment, you can act while the pendulum is still swinging. You can give it a push here, a shove there, trying to steer it toward the magnet you want. This is what metis feels like from the inside. You watch the system with utmost attention. You intervene where you can. It may work or it may not. But after years of doing it, you get a feel for which pushes help and which make things worse. You get better, but you never get certainty.

***

Scott writes: “We might reasonably think of [metis] as being partisan knowledge as opposed to generic knowledge. That is, the holder of such knowledge typically has a passionate interest in a particular outcome.“

An insurer of commercial shipping, he says, can afford to rely on probability distributions for accidents. But for a sailor hoping for a safe voyage, it is the outcome of a single trip that matters.

A general preparing for an attack uses the rule of thumb that the attacker needs a three-to-one numerical advantage. For the general, this ratio is what matters, not the outcome of any particular fight. But for the soldier on the ground, it’s the opposite. The outcome of the fight decides whether he lives to see the victory.

And so it is with our examples. Machiavelli struggles passionately to keep Florence alive. For him, it’s not a statistical exercise. The SRE is on the line for company profits. The farmers need their children fed. These are not hobbyists enjoying the thrill of uncertainty. They are people whose lives and livelihoods depend on getting the next call right, and all of them would rather not have to make it. When the stakes are high, making decisions under uncertainty isn’t exciting. It’s terrifying.

And all of them are trying to solve the problem not by acquiring more metis, but by circumventing it altogether. Machiavelli tries to chart a science of politics and diplomacy. The SRE introduces mechanical, automatically executed rules to deal with failures that once required human judgment. The farmer couple opts out entirely. They want their children to land in the safety and predictability of government jobs.

***

It's understandable. When faced with a problem that you cannot solve in full, when the best you can get is a heuristic, a rule of thumb, a gut feeling, then sooner or later you get tired of squinting at woolly bear caterpillars and start thinking about how to reshape the environment so that it becomes more predictable.

You put your turbulence-navigating skills aside for a moment and ask a different question: how do I move the system into one of those large monochromatic patches, where the pendulum's destination is largely predetermined, even in the face of large perturbations?

It can mean picking a more predictable job, or automating tasks. Or it can mean that you start planting trees in rectangular grids. And yes, a Prussian-style forest plantation looks terrible, but it can hardly be disputed that it surpasses the natural forest in predictability.

*** This may be the most charitable explanation of the high modernist impulse. Scott, all too often, makes it seem that it’s all about elites imposing legibility on their subjects. What he leaves largely unsaid is how the subjects themselves find metis exhausting and desperately want to escape it.

Proponents of microfinance love to stress how the poor are natural entrepreneurs. And entrepreneurship, of course, is the most metis-driven occupation imaginable. But as Banerjee and Duflo show in Poor Economics, the poor mostly do it because they have no other choice. Given the option, they would overwhelmingly prefer a salaried job and a predictable income.

So in a way, the subjects and the elites cooperate in the high modernist project, each for their own selfish reasons. Whether the outcome is forced collectivization or the Green Revolution depends very much on political details that have nothing to do with metis or the oh-so-common elite fetish for regular, rectangular grids.